stable diffusion
On Wednesday, the Future of Life Institute published an open letter on its website asking the AI lab to “stop training AI systems more powerful than GPT-4 for at least six months.” The letter, signed by Elon Musk and several prominent AI researchers, quickly began to attract press attention. criticism on social media.
Earlier this month, OpenAI released GPT-4. GPT-4 is an AI model that is said to be able to perform synthetic tasks and pass standardized tests at the human level, but these claims are still being evaluated by research. In any case, GPT-4 and Bing Chat’s advances in capabilities over previous AI models have surprised some experts who believe we are heading towards hyper-intelligent AI systems sooner than previously expected. I was.
Along these lines, the Future of Life Institute argues that recent advances in AI have led to an “out of control race” over the development and deployment of AI models that are difficult to predict and control. They believe that the lack of planning and management of these AI systems is a problem, and that powerful AI systems should only be developed after their effects are well understood and manageable. As they wrote in their letter:
AI systems with intelligence that competes with humans can pose serious risks to society and humanity, as shown by extensive research and recognized by leading AI labs. As stated in the widely-advocated Asilomar AI Principles, advanced AI has the potential to profoundly change the history of life on Earth and should be planned and managed with due care and resources.
In particular, the letter poses four questions. Some of them envision hypothetical scenarios that have been highly controversial in some parts of the AI community, such as AI losing “all jobs” and civilization’s “loss of control.”
- “should do it Are machines flooding our information channels with propaganda and lies? ”
- “should do it Do you want to automate all your work, including fulfillment work?“
- “should do it Will we eventually outnumber, become wiser, become obsolete, and develop non-human minds that may replace us?”
- “should do it Are you going to risk losing control of our civilization? ”
To address these potential threats, the letter calls on the AI Lab to “immediately suspend the training of AI systems stronger than GPT-4 for at least six months.” During the pause, the authors suggest that AI labs and independent experts work together to establish shared safety protocols for AI design and development. These protocols must be overseen by independent outside experts to ensure that AI systems are “safe beyond reasonable doubt.”
However, it is unclear what the phrase “more powerful than GPT-4” really means in a practical or regulatory sense. The letter does not specify how compliance is ensured by measuring the relative power of multimodal or large language models. Additionally, OpenAI explicitly avoids disclosing technical details about how GPT-4 works.
The Future of Life Institute is a non-profit organization founded in 2014 by a group of scientists concerned about the existential risks facing humanity, including biotechnology, nuclear weapons, and climate change. Additionally, hypothetical existential risk from AI is a key focus for this group. According to Reuters, the organization is primarily funded by the Musk Foundation, the Founders Pledge, an effective altruistic group based in London, and the Silicon Valley Community Foundation.
prominent signatories of the letter Confirmed by a Reuters reporter These include the aforementioned Tesla CEO Elon Musk, AI pioneers Joshua Bengio and Stuart Russell, Apple co-founder Steve Wozniak, Stability AI CEO Emad Mostok, and the author. Yuval Noah HarariAn open letter can be signed by anyone on the Internet without verification. This initially included incorrectly added names such as former Microsoft CEO Bill Gates, OpenAI CEO Sam Altman, and fictional characters. John WickThese names were later removed.
